DESCRIPTION (Applicant's abstract): This revised R01 application will examine the links between personal illness models and depression treatment adherence in primary care patients using state-of-the-art medication monitoring techniques. It will also explore their relationship to symptomatic and functional outcomes. By examining these complex relationships naturalistically, this research will lay the foundation for the design of intervention strategies to promote adherence to depression treatment in primary care practice. We propose to examine, over a 12-month period, the extent of nonadherence to pharmacologic treatment for depression among primary care patients, and to use Levanthal's self-regulatory model of illness cognition to evaluate the influence of these beliefs or representations of depressive illness on medication adherence. This study uses a longitudinal design to assess changes in these perceptions and treatment adherence over time Two hundred and ten primary care patients age 18 and older will be recruited from two family practice health centers and a network of eight family practice offices. Potential participants will be referred by their primary care physician Physicians will be asked to refer patients who have initiated antidepressant treatment within the preceding two weeks. Patients providing informed consent will complete five psychosocial assessments in a 12-month period (i.e., baseline, and months 3, 6, 9, and 12). Medication adherence will be assessed at these visits by patient self-report and electronic pill monitors. The proposed study will provide important information on patterns of nonadherence to antidepressant treatment the primary care sector during both acute and continuation phases of pharmacotherapy. It will also be the first study to examine illness representations for depression and their relationship to treatment adherence, and to assess change in personal illness models over time. The findings from this study will form the basis for development of adherence interventions, which are tailored to patients' self-management needs. Subsequent studies will be designed to evaluate the effectiveness of such interventions in the primary care sector.